V23G-06:
Mantle in the Manihiki Plateau source with ultra-depleted incompatible element abundances but FOZO-like isotopic signature
Tuesday, 16 December 2014: 2:55 PM
Roman Golowin1, Kaj Hoernle1, Maxim Portnyagin1, Folkmar Hauff1, Andrey Gurenko2, C-Dieter Garbe-Schoenberg3 and Reinhard Werner1, (1)GEOMAR Helmholtz Centre for Ocean Research Kiel, Kiel, Germany, (2)CRPG Centre de Recherches Pétrographiques et Géochimiques, Vandoeuvre-Les-Nancy, France, (3)University of Kiel, Institute of Geosciences, Kiel, Germany
Abstract:
The ~120Ma Manihiki Plateau basement consists of high-Ti tholeiitic basalts with EM-I type isotopic signatures, similar to the Singgalo basalts at Ontong Java, and low-Ti tholeiitic basalts with FOZO (Kwaimbaita/Kroenke) to HIMU-type isotopic compositions, similar to late stage volcanism on Hikurangi and Manihiki Plateaus (Hoernle et al. 2010; Timm et al. 2011). The low-Ti basalts have affinities to boninites and have been interpreted to be derived from residual mantle wedge mantle (Ingle et al. 2007). New major, volatile and trace element and radiogenic isotope data have been generated from fresh low-Ti glass samples recovered during R/V Sonne cruises SO193 and SO225. The low-Ti samples have distinctly lower Ti/V ratios compared to lavas from Ontong Java Plateau (Kwaimbaita-Kroenke and Singgalo), but similar to boninitic rocks. Glasses and melt inclusions in olivine have low volatile contents (0.12-0.25 wt% H2O). Olivine chemistry points to derivation from peridotite source. Therefore we interpret the low-Ti lavas to have formed through melting of dry and depleted peridotite at high temperatures, consistent with Timm et al (2011). The low-Ti group is characterized by U-shaped trace element patterns. The glass samples form linear mixing arrays on radiogenic isotope diagrams, pointing to the involvement of two components: 1) a component ultra-depleted in highly incompatible elements (UDC) but with intermediate Pb, Sr and Nd isotopic compositions, being similar to Kwaimbaita/Kroenke lavas from Ontong Java, and 2) an enriched component with HIMU-type incompatible element and isotopic characteristics, similar to late-stage volcanism on Manihiki, Hikurangi and Ontong Java (e.g. Hoernle et al. 2010). The ultra-depleted, FOZO-like mantle component could represent second stage melting of FOZO type mantle or re-melting of young recycled oceanic lithosphere within the plume head. Enrichment with HIMU type melts is required to explain the enrichment in the most incompatible elements and distinct U-shaped patterns of these melts. The presence of dry, boninite-like second stage melts in the Manihiki Plateau basement shows direct geochemical/petrological evidence for anomalously high temperatures in the upper mantle caused by a mantle plume.
